A note on the neoclassical ambiguity and the specific factor production model under variable returns to scale